Boston University
Algorithmic reasoning for multi-agent systems with limited information
Abstract
dc:description.abstractMulti-agent autonomous systems are becoming more prevalent than ever before, in warehouses and factories, on public roadways and sidewalks, for infrastructure inspection and ecological monitoring; The ability for autonomous agents to understand their surroundings, coordinate effectively, and navigate safely while completing their team objectives is paramount to successful operations. In well controlled environments, information, communication, and coordination are abundant, however, in real world deployments, these conveniences are few and far between. This dissertation explores a variety of topics that enhance the autonomy of individual and multi-agent systems through algorithmic reasoning. In this work, we present a variety of systematic methods to improve spatial awareness, high-level coordination, and safe operation through intuitive policies built from multi-objective optimization, environmental decomposition, and efficient graph algorithms. First, we explore how multi-objective optimization can be used to adaptively plan in highly dynamic and dangerous environments purely from local observations. Second, we investigate methods for generating Free-Space Ellipsoid Graphs in unknown and complex, non-convex environments from minimal and realistic sensor readings, which are used to encode spatial information in a discrete representation for the coordination of multiple heterogeneous agents. Finally, we build on our foundations of multi-objective optimization and network theory to develop minimally invasive observers that can extract a great deal of information from team operations. With these models, we investigate both how we can maintain team level privacy to protect sensitive objective information and how understanding team behaviors can be used to improve the coexistence between agents operating in shared and public environments.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy
- Level thesis:degree_level
- doctoral
- Discipline thesis:degree_discipline
- Mechanical Engineering
- Grantor
- Boston University
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Brodt, Brennan
- Advisor dc:contributor.advisor
-
- Pierson, Alyssa
- ORCID dc:identifier.orcid
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- Attribution 4.0 International
- Licence dc:rights.uri
- Language dc:language.iso
- en_US
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/2144/53028